Hair Relaxer Use and Risk of Uterine Leiomyomata in African-American Women

Lauren A. Wise; Julie R. Palmer; David Reich; Yvette C. Cozier; Lynn Rosenberg


Am J Epidemiol. 2012;175(5):432-440. 

In This Article

Materials and Methods

Study Population

The Black Women's Health Study is an ongoing US prospective cohort study of 59,000 African-American women aged 21–69 years at entry.[36] In 1995, Essence magazine subscribers, black members of 2 professional societies, and friends and relatives of early respondents were mailed an invitation to enroll in a long-term health study by completing a comprehensive self-administered baseline questionnaire. Every 2 years, participants are asked to update exposure and medical histories by questionnaire. Study participants reside in more than 17 states, with the majority residing in New York, California, Illinois, Michigan, Georgia, and New Jersey. The Institutional Review Board of Boston University Medical Center approved the study protocol.

Assessment of Outcome

On the 1999 and 2001 follow-up questionnaires, women reported whether they had been diagnosed with "uterine fibroids" in the previous 2-year interval, the calendar year in which they were first diagnosed, and whether their diagnosis was confirmed by "pelvic exam" and/or by "ultrasound/hysterectomy." On the 2003, 2005, 2007, and 2009 follow-up questionnaires, "hysterectomy" was replaced by "surgery (e.g., hysterectomy)" to capture women who had other surgeries such as myomectomy, and "ultrasound" and "surgery" were divided into separate questions. Cases were classified as "surgically-confirmed" if they reported diagnosis by "surgery" on the 2003 or later questionnaires, or if they reported a diagnosis by "ultrasound/hysterectomy" and also reported "hysterectomy" under a separate question in 1999 or 2001.

We included cases diagnosed by ultrasound as well as surgery because the latter represent only 10%–30% of diagnosed cases and because studies of such cases may spuriously identify risk factors associated with severity or treatment preference.[37] Ultrasound has high sensitivity (99%) and specificity (91%) relative to histologic evidence.[38,39] To maximize the specificity of uterine leiomyomata classification, pelvic examination cases (n = 538) were treated as noncases because these diagnoses could have represented other gynecologic pathology.[40]

Assessment of Hair Relaxer Use and Covariates

The baseline 1995 questionnaire collected data on several risk factors for uterine leiomyomata, including reproductive and contraceptive history; anthropometric factors; lifestyle factors (smoking, alcohol, physical activity); geographic region of residence; socioeconomic correlates (education, marital status, occupation); medical history; and gynecologic surveillance (recency of cervical smear and pelvic examination). In 1997, women reported their country of birth. Biennial follow-up questionnaires updated information on risk factors and identified new cases of uterine leiomyomata. Reproductive factors, weight, smoking, marital status, physical activity, and region were updated on follow-up questionnaires and were modeled as time-varying covariates in the analyses.

During 2002–2007, 50% of study participants provided a mouthwash-saliva sample as a source of DNA.[41] As part of a genetic admixture mapping study in our cohort,[42–44] we used a set of ancestral informative genetic polymorphisms to estimate percentage of European ancestry for 2,259 uterine leiomyomata cases and 2,154 unaffected women who had provided a DNA sample. Average "European ancestry" among African Americans is estimated to be 20%.[45]

The 1997 follow-up questionnaire included questions about use of "chemical hair straighteners": age at first use (<10, 10–19, 20–29, ≥30 years); frequency of use per year (about 1 time, 2 times, 3–4 times, 5–6 times, ≥7 times); total number of years of use (<1, 1–4, 5–9, 10–14, 15–19, ≥20); number of burns (a break in the skin, not just tingling) during the application of chemical straighteners (never, 1–2, 3–4, 5–9, ≥10 times); and type of formulation used most often (lye, no-lye, don't know).

Validation Studies

With regard to uterine leiomyomata, a random sample of cases (n = 248) were mailed supplemental questionnaires regarding their initial date of diagnosis, method(s) of confirmation, symptoms, and treatment, and they were asked for permission to review their medical records. We obtained medical records from 127 of the 128 women who gave us permission and confirmed the self-report in 122 (96%). Among the 188 (76%) who provided supplemental survey data, 71% reported uterine leiomyomata-related symptoms prior to diagnosis, and 87% reported that their condition came to clinical attention because they sought treatment for symptoms or a tumor was palpable during a routine pelvic examination. There were no appreciable differences between cases who did and did not release medical records with respect to uterine leiomyomata risk factors.[46]

Restriction Criteria

Follow-up began in 1997 because the method of uterine leiomyomata diagnosis was first included on the 1999 questionnaire. Of the 53,126 respondents to the 1997 questionnaire, we excluded postmenopausal women (n = 16,478) in whom uterine leiomyomata are rare;[47] women with a history of uterine leiomyomata (n = 10,655); women who did not complete a questionnaire after 1997 (n = 940); cases without data on diagnosis year (n = 120) or method (n = 112); and women with missing exposure (n = 659) or covariate data (n = 582), leaving 23,580 women for analysis. Those excluded had lower educational attainment than those included but were similar with respect to parity, age at menarche, and other established uterine leiomyomata risk factors.

Data Analysis

We defined cases as women who reported a first diagnosis of uterine leiomyomata confirmed by ultrasound or surgery. Person-years were calculated from March 1997 until uterine leiomyomata diagnosis (n = 7,146), menopause (n = 2,639), death (n = 372), loss to follow-up (n = 3,393), or March 2009 (n = 10,030), whichever came first. Age- and period-stratified Cox regression was used to estimate incidence rate ratios and 95% confidence intervals for the associations of interest.

Hair relaxer variables were categorized according to their frequency distributions in the analytical sample. We also created a hair relaxer score variable, the product of frequency and duration of use, after first converting them to continuous variables by assigning values to the midpoint of each category. We defined "never users" as those who reported never using hair relaxers or having used them for less than 1 year in total. A covariate was included in multivariable analyses if it was either an established risk factor for uterine leiomyomata identified from the literature or a potential risk factor for uterine leiomyomata associated with hair relaxer use. On the basis of these criteria, we constructed a multivariable model that controlled for age (1-year intervals), time period (2-year intervals), age at menarche (years), parity (births), age at first birth (years), years since last birth (<5, 5–9, 10–14, 15, ≥20), oral contraceptive use (never, current, past), body mass index (<20, 20–24, 25–29, 30–34, ≥35 kg/m2), smoking (current, past, never), current alcohol use (<1, 1–6, ≥7 drinks/week), education (≤12, 13–15, 16, ≥17 years), marital status (married/partnered, divorced/separated/widowed, single), occupation (white collar, non-white collar, unemployed, missing), household income (≤$25,000, $25,001–$50,000, $50,001–$100,000, >$100,000, missing), geographic region (South, Northeast, Midwest, and West), and country of birth (United States vs. outside United States).

Tests for trend were conducted by modeling the ordinal categorical version of the exposure variable among the ever users of hair relaxers.[48] We used stratification methods to examine whether main associations were modified by other covariates such as body mass index, education, and geographic region. P values from interaction tests were obtained by using the likelihood ratio test comparing models with and without cross-product terms between the covariate and exposure variable. Departures from proportional hazards were evaluated in the same manner by using cross-product terms between each exposure variable and age (<35 vs. ≥35 years) and time period (1997–2003 vs. 2003–2009). Analyses were performed by using SAS, version 9.2, statistical software.[49]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: